* kernel 3.5.2/amd64: iwlwifi 0000:03:00.0: failed to allocate pci memory @ 2012-09-04 0:30 ` Marc MERLIN 2012-09-08 17:01 ` Marc MERLIN 0 siblings, 1 reply; 8+ messages in thread From: Marc MERLIN @ 2012-09-04 0:30 UTC (permalink / raw) To: linux-wireless Howdy, I currently rmmod iwlwifi before putting my laptop to sleep and reload it when coming back. Arguably, it's maybe not needed, but from time to time I hit this memory allocation failure below: My kenrel config options are here http://marc.merlins.org/tmp/config-3.5.2-amd64-preempt-noide-20120731 and this happens on an Lenovo T530. dmesg: [856806.443647] cfg80211: Calling CRDA to update world regulatory domain [856806.448428] iwlwifi: Intel(R) Wireless WiFi Link AGN driver for Linux, in-tree:d [856806.448431] iwlwifi: Copyright(c) 2003-2012 Intel Corporation [856806.448929] cfg80211: World regulatory domain updated: [856806.448931] cfg80211: (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp) [856806.448933] cfg80211: (2402000 KHz - 2472000 KHz @ 40000 KHz), (300 mBi, 2000 mBm) [856806.448941] cfg80211: (2457000 KHz - 2482000 KHz @ 20000 KHz), (300 mBi, 2000 mBm) [856806.448942] cfg80211: (2474000 KHz - 2494000 KHz @ 20000 KHz), (300 mBi, 2000 mBm) [856806.448943] cfg80211: (5170000 KHz - 5250000 KHz @ 40000 KHz), (300 mBi, 2000 mBm) [856806.448945] cfg80211: (5735000 KHz - 5835000 KHz @ 40000 KHz), (300 mBi, 2000 mBm) [856806.483929] iwlwifi 0000:03:00.0: pci_resource_len = 0x00002000 [856806.483932] iwlwifi 0000:03:00.0: pci_resource_base = ffffc900057bc000 [856806.483933] iwlwifi 0000:03:00.0: HW Revision ID = 0x3E [856806.484004] iwlwifi 0000:03:00.0: irq 46 for MSI/MSI-X [856806.497476] iwlwifi 0000:03:00.0: loaded firmware version 9.221.4.1 build 25532 [856806.497944] kworker/3:0: page allocation failure: order:5, mode:0xd0 [856806.497948] Pid: 17936, comm: kworker/3:0 Tainted: G W O 3.5.2-amd64-preempt-noide-20120731 #1 [856806.497949] Call Trace: [856806.497959] [<ffffffff810cf54c>] warn_alloc_failed+0x117/0x12c [856806.497963] [<ffffffff810d23af>] __alloc_pages_nodemask+0x6e3/0x792 [856806.497969] [<ffffffff812b7f41>] ? pfn_to_dma_pte+0x116/0x15e [856806.497976] [<ffffffff810ff58b>] alloc_pages_current+0xcd/0xee [856806.497979] [<ffffffff810cecca>] __get_free_pages+0x9/0x45 [856806.497982] [<ffffffff812ba67d>] intel_alloc_coherent+0x84/0xe7 [856806.497986] [<ffffffff81085cf8>] ? arch_local_irq_save+0x15/0x1b [856806.497999] [<ffffffffa0b84afc>] iwl_ucode_callback+0xa49/0xc0d [iwlwifi] [856806.498006] [<ffffffff8128f100>] ? _request_firmware_prepare.isra.5+0x1bf/0x1bf [856806.498010] [<ffffffff8128f181>] request_firmware_work_func+0x81/0xb1 [856806.498014] [<ffffffff81054b13>] process_one_work+0x16f/0x28e [856806.498018] [<ffffffff810555d5>] worker_thread+0xce/0x152 [856806.498021] [<ffffffff81055507>] ? manage_workers.isra.24+0x16c/0x16c [856806.498024] [<ffffffff81058e3c>] kthread+0x86/0x8e [856806.498029] [<ffffffff813a0aa4>] kernel_thread_helper+0x4/0x10 [856806.498032] [<ffffffff81058db6>] ? kthread_freezable_should_stop+0x3e/0x3e [856806.498034] [<ffffffff813a0aa0>] ? gs_change+0x13/0x13 [856806.498036] Mem-Info: [856806.498037] Node 0 DMA per-cpu: [856806.498039] CPU 0: hi: 0, btch: 1 usd: 0 [856806.498041] CPU 1: hi: 0, btch: 1 usd: 0 [856806.498042] CPU 2: hi: 0, btch: 1 usd: 0 [856806.498044] CPU 3: hi: 0, btch: 1 usd: 0 [856806.498045] Node 0 DMA32 per-cpu: [856806.498047] CPU 0: hi: 186, btch: 31 usd: 0 [856806.498048] CPU 1: hi: 186, btch: 31 usd: 0 [856806.498050] CPU 2: hi: 186, btch: 31 usd: 0 [856806.498051] CPU 3: hi: 186, btch: 31 usd: 0 [856806.498052] Node 0 Normal per-cpu: [856806.498054] CPU 0: hi: 186, btch: 31 usd: 0 [856806.498055] CPU 1: hi: 186, btch: 31 usd: 0 [856806.498057] CPU 2: hi: 186, btch: 31 usd: 0 [856806.498058] CPU 3: hi: 186, btch: 31 usd: 0 [856806.498062] active_anon:880341 inactive_anon:274439 isolated_anon:0 [856806.498062] active_file:222778 inactive_file:228271 isolated_file:0 [856806.498062] unevictable:1436 dirty:369 writeback:0 unstable:0 [856806.498062] free:137592 slab_reclaimable:167927 slab_unreclaimable:17618 [856806.498062] mapped:27053 shmem:29116 pagetables:23832 bounce:0 [856806.498065] Node 0 DMA free:15900kB min:132kB low:164kB high:196kB active_anon:0kB inactive_anon:0kB active_file:0kB inactive_file:0kB unevictable:0kB isolated(anon):0kB isolated(file):0kB present:15676kB mlocked:0kB dirty:0kB writeback:0kB mapped:0kB shmem:0kB slab_reclaimable:0kB slab_unreclaimable:0kB kernel_stack:0kB pagetables:0kB unstable:0kB bounce:0kB writeback_tmp:0kB pages_scanned:0 all_unreclaimable? yes [856806.498071] lowmem_reserve[]: 0 3257 7777 7777 [856806.498074] Node 0 DMA32 free:454560kB min:28252kB low:35312kB high:42376kB active_anon:1412512kB inactive_anon:513084kB active_file:333072kB inactive_file:323580kB unevictable:196kB isolated(anon):0kB isolated(file):0kB present:3335900kB mlocked:196kB dirty:108kB writeback:0kB mapped:43284kB shmem:27472kB slab_reclaimable:281024kB slab_unreclaimable:20952kB kernel_stack:4200kB pagetables:27640kB unstable:0kB bounce:0kB writeback_tmp:0kB pages_scanned:0 all_unreclaimable? no [856806.498079] lowmem_reserve[]: 0 0 4519 4519 [856806.498083] Node 0 Normal free:79908kB min:39196kB low:48992kB high:58792kB active_anon:2108852kB inactive_anon:584672kB active_file:558040kB inactive_file:589504kB unevictable:5548kB isolated(anon):0kB isolated(file):0kB present:4627820kB mlocked:5548kB dirty:1368kB writeback:0kB mapped:64928kB shmem:88992kB slab_reclaimable:390684kB slab_unreclaimable:49520kB kernel_stack:3496kB pagetables:67688kB unstable:0kB bounce:0kB writeback_tmp:0kB pages_scanned:0 all_unreclaimable? no [856806.498088] lowmem_reserve[]: 0 0 0 0 [856806.498090] Node 0 DMA: 1*4kB 1*8kB 1*16kB 0*32kB 2*64kB 1*128kB 1*256kB 0*512kB 1*1024kB 1*2048kB 3*4096kB = 15900kB [856806.498098] Node 0 DMA32: 60562*4kB 25568*8kB 264*16kB 4*32kB 1*64kB 1*128kB 0*256kB 2*512kB 2*1024kB 0*2048kB 0*4096kB = 454408kB [856806.498106] Node 0 Normal: 15497*4kB 861*8kB 162*16kB 79*32kB 22*64kB 4*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 1*4096kB = 80012kB [856806.498115] 660922 total pagecache pages [856806.498116] 180029 pages in swap cache [856806.498118] Swap cache stats: add 7904087, delete 7724058, find 2010810/2392461 [856806.498119] Free swap = 5831292kB [856806.498121] Total swap = 10485756kB [856806.520762] 2057712 pages RAM [856806.520764] 63301 pages reserved [856806.520765] 543630 pages shared [856806.520766] 1544767 pages non-shared [856806.520771] iwlwifi 0000:03:00.0: failed to allocate pci memory It will typically fix itself and work again later. Any ideas? Thanks, Marc -- "A mouse is a device used to point at the xterm you want to type in" - A.S.R. Microsoft is to operating systems .... .... what McDonalds is to gourmet cooking Home page: http://marc.merlins.org/ ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: kernel 3.5.2/amd64: iwlwifi 0000:03:00.0: failed to allocate pci memory 2012-09-04 0:30 ` kernel 3.5.2/amd64: iwlwifi 0000:03:00.0: failed to allocate pci memory Marc MERLIN @ 2012-09-08 17:01 ` Marc MERLIN 2012-09-08 18:57 ` Johannes Berg 0 siblings, 1 reply; 8+ messages in thread From: Marc MERLIN @ 2012-09-08 17:01 UTC (permalink / raw) To: johannes.berg, wey-yi.w.guy, ilw; +Cc: linux-wireless, netdev Howdy, I currently rmmod iwlwifi before putting my laptop to sleep and reload it when coming back. Arguably, it's maybe not needed, but from time to time I hit this memory allocation failure below. I realize it's likely a memory fragmentation problem, but I have 8GB and plenty of 'free' space, so I'm hoping that somehow it can be defragmented enough for module loading ot work? My kenrel config options are here http://marc.merlins.org/tmp/config-3.5.2-amd64-preempt-noide-20120731 and this happens on an Lenovo T530 when it works: [ 13.494270] iwlwifi 0000:03:00.0: loaded firmware version 9.221.4.1 build 25532 [ 13.494440] iwlwifi 0000:03:00.0: Detected Intel(R) Centrino(R) Ultimate-N 6300 AGN, REV=0x74 When it doesn't: [856806.443647] cfg80211: Calling CRDA to update world regulatory domain [856806.448428] iwlwifi: Intel(R) Wireless WiFi Link AGN driver for Linux, in-tree:d [856806.448431] iwlwifi: Copyright(c) 2003-2012 Intel Corporation [856806.448929] cfg80211: World regulatory domain updated: [856806.448931] cfg80211: (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp) [856806.448933] cfg80211: (2402000 KHz - 2472000 KHz @ 40000 KHz), (300 mBi, 2000 mBm) [856806.448941] cfg80211: (2457000 KHz - 2482000 KHz @ 20000 KHz), (300 mBi, 2000 mBm) [856806.448942] cfg80211: (2474000 KHz - 2494000 KHz @ 20000 KHz), (300 mBi, 2000 mBm) [856806.448943] cfg80211: (5170000 KHz - 5250000 KHz @ 40000 KHz), (300 mBi, 2000 mBm) [856806.448945] cfg80211: (5735000 KHz - 5835000 KHz @ 40000 KHz), (300 mBi, 2000 mBm) [856806.483929] iwlwifi 0000:03:00.0: pci_resource_len = 0x00002000 [856806.483932] iwlwifi 0000:03:00.0: pci_resource_base = ffffc900057bc000 [856806.483933] iwlwifi 0000:03:00.0: HW Revision ID = 0x3E [856806.484004] iwlwifi 0000:03:00.0: irq 46 for MSI/MSI-X [856806.497476] iwlwifi 0000:03:00.0: loaded firmware version 9.221.4.1 build 25532 [856806.497944] kworker/3:0: page allocation failure: order:5, mode:0xd0 [856806.497948] Pid: 17936, comm: kworker/3:0 Tainted: G W O 3.5.2-amd64-preempt-noide-20120731 #1 [856806.497949] Call Trace: [856806.497959] [<ffffffff810cf54c>] warn_alloc_failed+0x117/0x12c [856806.497963] [<ffffffff810d23af>] __alloc_pages_nodemask+0x6e3/0x792 [856806.497969] [<ffffffff812b7f41>] ? pfn_to_dma_pte+0x116/0x15e [856806.497976] [<ffffffff810ff58b>] alloc_pages_current+0xcd/0xee [856806.497979] [<ffffffff810cecca>] __get_free_pages+0x9/0x45 [856806.497982] [<ffffffff812ba67d>] intel_alloc_coherent+0x84/0xe7 [856806.497986] [<ffffffff81085cf8>] ? arch_local_irq_save+0x15/0x1b [856806.497999] [<ffffffffa0b84afc>] iwl_ucode_callback+0xa49/0xc0d [iwlwifi] [856806.498006] [<ffffffff8128f100>] ? _request_firmware_prepare.isra.5+0x1bf/0x1bf [856806.498010] [<ffffffff8128f181>] request_firmware_work_func+0x81/0xb1 [856806.498014] [<ffffffff81054b13>] process_one_work+0x16f/0x28e [856806.498018] [<ffffffff810555d5>] worker_thread+0xce/0x152 [856806.498021] [<ffffffff81055507>] ? manage_workers.isra.24+0x16c/0x16c [856806.498024] [<ffffffff81058e3c>] kthread+0x86/0x8e [856806.498029] [<ffffffff813a0aa4>] kernel_thread_helper+0x4/0x10 [856806.498032] [<ffffffff81058db6>] ? kthread_freezable_should_stop+0x3e/0x3e [856806.498034] [<ffffffff813a0aa0>] ? gs_change+0x13/0x13 [856806.498036] Mem-Info: [856806.498037] Node 0 DMA per-cpu: [856806.498039] CPU 0: hi: 0, btch: 1 usd: 0 [856806.498041] CPU 1: hi: 0, btch: 1 usd: 0 [856806.498042] CPU 2: hi: 0, btch: 1 usd: 0 [856806.498044] CPU 3: hi: 0, btch: 1 usd: 0 [856806.498045] Node 0 DMA32 per-cpu: [856806.498047] CPU 0: hi: 186, btch: 31 usd: 0 [856806.498048] CPU 1: hi: 186, btch: 31 usd: 0 [856806.498050] CPU 2: hi: 186, btch: 31 usd: 0 [856806.498051] CPU 3: hi: 186, btch: 31 usd: 0 [856806.498052] Node 0 Normal per-cpu: [856806.498054] CPU 0: hi: 186, btch: 31 usd: 0 [856806.498055] CPU 1: hi: 186, btch: 31 usd: 0 [856806.498057] CPU 2: hi: 186, btch: 31 usd: 0 [856806.498058] CPU 3: hi: 186, btch: 31 usd: 0 [856806.498062] active_anon:880341 inactive_anon:274439 isolated_anon:0 [856806.498062] active_file:222778 inactive_file:228271 isolated_file:0 [856806.498062] unevictable:1436 dirty:369 writeback:0 unstable:0 [856806.498062] free:137592 slab_reclaimable:167927 slab_unreclaimable:17618 [856806.498062] mapped:27053 shmem:29116 pagetables:23832 bounce:0 [856806.498065] Node 0 DMA free:15900kB min:132kB low:164kB high:196kB active_anon:0kB inactive_anon:0kB active_file:0kB inactive_file:0kB unevictable:0kB isolated(anon):0kB isolated(file):0kB present:15676kB mlocked:0kB dirty:0kB writeback:0kB mapped:0kB shmem:0kB slab_reclaimable:0kB slab_unreclaimable:0kB kernel_stack:0kB pagetables:0kB unstable:0kB bounce:0kB writeback_tmp:0kB pages_scanned:0 all_unreclaimable? yes [856806.498071] lowmem_reserve[]: 0 3257 7777 7777 [856806.498074] Node 0 DMA32 free:454560kB min:28252kB low:35312kB high:42376kB active_anon:1412512kB inactive_anon:513084kB active_file:333072kB inactive_file:323580kB unevictable:196kB isolated(anon):0kB isolated(file):0kB present:3335900kB mlocked:196kB dirty:108kB writeback:0kB mapped:43284kB shmem:27472kB slab_reclaimable:281024kB slab_unreclaimable:20952kB kernel_stack:4200kB pagetables:27640kB unstable:0kB bounce:0kB writeback_tmp:0kB pages_scanned:0 all_unreclaimable? no [856806.498079] lowmem_reserve[]: 0 0 4519 4519 [856806.498083] Node 0 Normal free:79908kB min:39196kB low:48992kB high:58792kB active_anon:2108852kB inactive_anon:584672kB active_file:558040kB inactive_file:589504kB unevictable:5548kB isolated(anon):0kB isolated(file):0kB present:4627820kB mlocked:5548kB dirty:1368kB writeback:0kB mapped:64928kB shmem:88992kB slab_reclaimable:390684kB slab_unreclaimable:49520kB kernel_stack:3496kB pagetables:67688kB unstable:0kB bounce:0kB writeback_tmp:0kB pages_scanned:0 all_unreclaimable? no [856806.498088] lowmem_reserve[]: 0 0 0 0 [856806.498090] Node 0 DMA: 1*4kB 1*8kB 1*16kB 0*32kB 2*64kB 1*128kB 1*256kB 0*512kB 1*1024kB 1*2048kB 3*4096kB = 15900kB [856806.498098] Node 0 DMA32: 60562*4kB 25568*8kB 264*16kB 4*32kB 1*64kB 1*128kB 0*256kB 2*512kB 2*1024kB 0*2048kB 0*4096kB = 454408kB [856806.498106] Node 0 Normal: 15497*4kB 861*8kB 162*16kB 79*32kB 22*64kB 4*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 1*4096kB = 80012kB [856806.498115] 660922 total pagecache pages [856806.498116] 180029 pages in swap cache [856806.498118] Swap cache stats: add 7904087, delete 7724058, find 2010810/2392461 [856806.498119] Free swap = 5831292kB [856806.498121] Total swap = 10485756kB [856806.520762] 2057712 pages RAM [856806.520764] 63301 pages reserved [856806.520765] 543630 pages shared [856806.520766] 1544767 pages non-shared [856806.520771] iwlwifi 0000:03:00.0: failed to allocate pci memory It will typically fix itself and work again later. Any ideas? Thanks, Marc -- "A mouse is a device used to point at the xterm you want to type in" - A.S.R. Microsoft is to operating systems .... .... what McDonalds is to gourmet cooking Home page: http://marc.merlins.org/ ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: kernel 3.5.2/amd64: iwlwifi 0000:03:00.0: failed to allocate pci memory 2012-09-08 17:01 ` Marc MERLIN @ 2012-09-08 18:57 ` Johannes Berg 2012-09-10 10:24 ` Johannes Berg 0 siblings, 1 reply; 8+ messages in thread From: Johannes Berg @ 2012-09-08 18:57 UTC (permalink / raw) To: Marc MERLIN; +Cc: wey-yi.w.guy, ilw, linux-wireless, netdev On Sat, 2012-09-08 at 10:01 -0700, Marc MERLIN wrote: > I realize it's likely a memory fragmentation problem, but I have 8GB and > plenty of 'free' space, so I'm hoping that somehow it can be defragmented > enough for module loading ot work? > [856806.443647] cfg80211: Calling CRDA to update world regulatory domain > [856806.448428] iwlwifi: Intel(R) Wireless WiFi Link AGN driver for Linux, in-tree:d > [856806.448431] iwlwifi: Copyright(c) 2003-2012 Intel Corporation > [856806.448929] cfg80211: World regulatory domain updated: > [856806.448931] cfg80211: (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp) > [856806.448933] cfg80211: (2402000 KHz - 2472000 KHz @ 40000 KHz), (300 mBi, 2000 mBm) > [856806.448941] cfg80211: (2457000 KHz - 2482000 KHz @ 20000 KHz), (300 mBi, 2000 mBm) > [856806.448942] cfg80211: (2474000 KHz - 2494000 KHz @ 20000 KHz), (300 mBi, 2000 mBm) > [856806.448943] cfg80211: (5170000 KHz - 5250000 KHz @ 40000 KHz), (300 mBi, 2000 mBm) > [856806.448945] cfg80211: (5735000 KHz - 5835000 KHz @ 40000 KHz), (300 mBi, 2000 mBm) > [856806.483929] iwlwifi 0000:03:00.0: pci_resource_len = 0x00002000 > [856806.483932] iwlwifi 0000:03:00.0: pci_resource_base = ffffc900057bc000 > [856806.483933] iwlwifi 0000:03:00.0: HW Revision ID = 0x3E > [856806.484004] iwlwifi 0000:03:00.0: irq 46 for MSI/MSI-X > [856806.497476] iwlwifi 0000:03:00.0: loaded firmware version 9.221.4.1 build 25532 > [856806.497944] kworker/3:0: page allocation failure: order:5, mode:0xd0 > [856806.497948] Pid: 17936, comm: kworker/3:0 Tainted: G W O 3.5.2-amd64-preempt-noide-20120731 #1 > [856806.497949] Call Trace: > [856806.497959] [<ffffffff810cf54c>] warn_alloc_failed+0x117/0x12c > [856806.497963] [<ffffffff810d23af>] __alloc_pages_nodemask+0x6e3/0x792 > [856806.497969] [<ffffffff812b7f41>] ? pfn_to_dma_pte+0x116/0x15e > [856806.497976] [<ffffffff810ff58b>] alloc_pages_current+0xcd/0xee > [856806.497979] [<ffffffff810cecca>] __get_free_pages+0x9/0x45 > [856806.497982] [<ffffffff812ba67d>] intel_alloc_coherent+0x84/0xe7 > [856806.497986] [<ffffffff81085cf8>] ? arch_local_irq_save+0x15/0x1b > [856806.497999] [<ffffffffa0b84afc>] iwl_ucode_callback+0xa49/0xc0d [iwlwifi] Yes, unfortunately we need a whole bunch of contiguous memory to load the firmware. > Any ideas? Nothing we can do from the driver side, I'm afraid. johannes ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: kernel 3.5.2/amd64: iwlwifi 0000:03:00.0: failed to allocate pci memory 2012-09-08 18:57 ` Johannes Berg @ 2012-09-10 10:24 ` Johannes Berg 2012-09-10 15:12 ` Larry Finger 2012-09-10 15:34 ` Marc MERLIN 0 siblings, 2 replies; 8+ messages in thread From: Johannes Berg @ 2012-09-10 10:24 UTC (permalink / raw) To: Marc MERLIN; +Cc: wey-yi.w.guy, ilw, linux-wireless, netdev On Sat, 2012-09-08 at 20:57 +0200, Johannes Berg wrote: > > [856806.497959] [<ffffffff810cf54c>] warn_alloc_failed+0x117/0x12c > > [856806.497963] [<ffffffff810d23af>] __alloc_pages_nodemask+0x6e3/0x792 > > [856806.497969] [<ffffffff812b7f41>] ? pfn_to_dma_pte+0x116/0x15e > > [856806.497976] [<ffffffff810ff58b>] alloc_pages_current+0xcd/0xee > > [856806.497979] [<ffffffff810cecca>] __get_free_pages+0x9/0x45 > > [856806.497982] [<ffffffff812ba67d>] intel_alloc_coherent+0x84/0xe7 > > [856806.497986] [<ffffffff81085cf8>] ? arch_local_irq_save+0x15/0x1b > > [856806.497999] [<ffffffffa0b84afc>] iwl_ucode_callback+0xa49/0xc0d [iwlwifi] > > Yes, unfortunately we need a whole bunch of contiguous memory to load > the firmware. > > > Any ideas? > > Nothing we can do from the driver side, I'm afraid. Turns out I was wrong. Here's a patch you can test. Note that we still need a lot of DMA-coherent memory for other things, but at least for the firmware image we don't. http://p.sipsolutions.net/11ea33b376a5bac5.txt johannes ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: kernel 3.5.2/amd64: iwlwifi 0000:03:00.0: failed to allocate pci memory 2012-09-10 10:24 ` Johannes Berg @ 2012-09-10 15:12 ` Larry Finger 2012-09-10 15:44 ` Johannes Berg 2012-09-10 15:34 ` Marc MERLIN 1 sibling, 1 reply; 8+ messages in thread From: Larry Finger @ 2012-09-10 15:12 UTC (permalink / raw) To: Johannes Berg; +Cc: Marc MERLIN, wey-yi.w.guy, ilw, linux-wireless, netdev On 09/10/2012 05:24 AM, Johannes Berg wrote: > On Sat, 2012-09-08 at 20:57 +0200, Johannes Berg wrote: > >>> [856806.497959] [<ffffffff810cf54c>] warn_alloc_failed+0x117/0x12c >>> [856806.497963] [<ffffffff810d23af>] __alloc_pages_nodemask+0x6e3/0x792 >>> [856806.497969] [<ffffffff812b7f41>] ? pfn_to_dma_pte+0x116/0x15e >>> [856806.497976] [<ffffffff810ff58b>] alloc_pages_current+0xcd/0xee >>> [856806.497979] [<ffffffff810cecca>] __get_free_pages+0x9/0x45 >>> [856806.497982] [<ffffffff812ba67d>] intel_alloc_coherent+0x84/0xe7 >>> [856806.497986] [<ffffffff81085cf8>] ? arch_local_irq_save+0x15/0x1b >>> [856806.497999] [<ffffffffa0b84afc>] iwl_ucode_callback+0xa49/0xc0d [iwlwifi] >> >> Yes, unfortunately we need a whole bunch of contiguous memory to load >> the firmware. >> >>> Any ideas? >> >> Nothing we can do from the driver side, I'm afraid. > > Turns out I was wrong. Here's a patch you can test. Note that we still > need a lot of DMA-coherent memory for other things, but at least for the > firmware image we don't. > > http://p.sipsolutions.net/11ea33b376a5bac5.txt That patch looks like a good way to solve the problem; however, some architectures need an explicit "#include <linux/vmalloc.h>" somewhere in the headers, even though x86 does not, and I did not see it in the patch. I didn't do a lot of checking, but grep did not find one in the existing code. Larry ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: kernel 3.5.2/amd64: iwlwifi 0000:03:00.0: failed to allocate pci memory 2012-09-10 15:12 ` Larry Finger @ 2012-09-10 15:44 ` Johannes Berg 0 siblings, 0 replies; 8+ messages in thread From: Johannes Berg @ 2012-09-10 15:44 UTC (permalink / raw) To: Larry Finger; +Cc: Marc MERLIN, wey-yi.w.guy, ilw, linux-wireless On Mon, 2012-09-10 at 10:12 -0500, Larry Finger wrote: > > http://p.sipsolutions.net/11ea33b376a5bac5.txt > > That patch looks like a good way to solve the problem; however, some > architectures need an explicit "#include <linux/vmalloc.h>" somewhere in the > headers, even though x86 does not, and I did not see it in the patch. I didn't > do a lot of checking, but grep did not find one in the existing code. Thanks for that, I added the include. johannes ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: kernel 3.5.2/amd64: iwlwifi 0000:03:00.0: failed to allocate pci memory 2012-09-10 10:24 ` Johannes Berg 2012-09-10 15:12 ` Larry Finger @ 2012-09-10 15:34 ` Marc MERLIN [not found] ` <20120910154402.GF5053@merlins.org> 1 sibling, 1 reply; 8+ messages in thread From: Marc MERLIN @ 2012-09-10 15:34 UTC (permalink / raw) To: Johannes Berg; +Cc: wey-yi.w.guy, ilw, linux-wireless, netdev On Mon, Sep 10, 2012 at 12:24:01PM +0200, Johannes Berg wrote: > On Sat, 2012-09-08 at 20:57 +0200, Johannes Berg wrote: > > > > [856806.497959] [<ffffffff810cf54c>] warn_alloc_failed+0x117/0x12c > > > [856806.497963] [<ffffffff810d23af>] __alloc_pages_nodemask+0x6e3/0x792 > > > [856806.497969] [<ffffffff812b7f41>] ? pfn_to_dma_pte+0x116/0x15e > > > [856806.497976] [<ffffffff810ff58b>] alloc_pages_current+0xcd/0xee > > > [856806.497979] [<ffffffff810cecca>] __get_free_pages+0x9/0x45 > > > [856806.497982] [<ffffffff812ba67d>] intel_alloc_coherent+0x84/0xe7 > > > [856806.497986] [<ffffffff81085cf8>] ? arch_local_irq_save+0x15/0x1b > > > [856806.497999] [<ffffffffa0b84afc>] iwl_ucode_callback+0xa49/0xc0d [iwlwifi] > > > > Yes, unfortunately we need a whole bunch of contiguous memory to load > > the firmware. > > > > > Any ideas? > > > > Nothing we can do from the driver side, I'm afraid. > > Turns out I was wrong. Here's a patch you can test. Note that we still > need a lot of DMA-coherent memory for other things, but at least for the > firmware image we don't. Thanks for that, I'll try it out and report back if I have other problems (it's hard to prove that it worked since I don't have the fragmented memory right away). Marc -- "A mouse is a device used to point at the xterm you want to type in" - A.S.R. Microsoft is to operating systems .... .... what McDonalds is to gourmet cooking Home page: http://marc.merlins.org/ ^ permalink raw reply [flat|nested] 8+ messages in thread
[parent not found: <20120910154402.GF5053@merlins.org>]
[parent not found: <1347292023.4272.28.camel@jlt4.sipsolutions.net>]
* Re: kernel 3.5.2/amd64: iwlwifi 0000:03:00.0: failed to allocate pci memory [not found] ` <1347292023.4272.28.camel@jlt4.sipsolutions.net> @ 2012-09-13 4:52 ` Marc MERLIN 0 siblings, 0 replies; 8+ messages in thread From: Marc MERLIN @ 2012-09-13 4:52 UTC (permalink / raw) To: Johannes Berg; +Cc: wey-yi.w.guy, ilw, linux-wireless, netdev [-- Attachment #1: Type: text/plain, Size: 898 bytes --] On Mon, Sep 10, 2012 at 05:47:03PM +0200, Johannes Berg wrote: > > but http://p.sipsolutions.net/11ea33b376a5bac5.txt > > refers to drivers/net/wireless/iwlwifi/dvm/ucode.c > > > > Obviously I can fix pathnames by hand, but jus wanted to make sure I'm doing > > the right thing before doing that. > > That'll probably just work. I may also erroneously have included > dvm16/... changes that are internal only, so you won't have to worry > about that file. I had to port the patch back to my 3.5.3 kernel and since then I haven't hany firmware loading failures, thanks much. I'll attach the patch below in case it helps others. Thanks for your help, Marc -- "A mouse is a device used to point at the xterm you want to type in" - A.S.R. Microsoft is to operating systems .... .... what McDonalds is to gourmet cooking Home page: http://marc.merlins.org/ [-- Attachment #2: iwlwifi-firmware.patch --] [-- Type: text/plain, Size: 5010 bytes --] diff -urN .iwlwifi/iwl-drv.c iwlwifi/iwl-drv.c --- .iwlwifi/iwl-drv.c 2012-07-21 13:58:29.000000000 -0700 +++ iwlwifi/iwl-drv.c 2012-09-12 16:30:28.997944875 -0700 @@ -116,10 +116,8 @@ static void iwl_free_fw_desc(struct iwl_drv *drv, struct fw_desc *desc) { - if (desc->v_addr) - dma_free_coherent(drv->trans->dev, desc->len, - desc->v_addr, desc->p_addr); - desc->v_addr = NULL; + vfree(desc->data); + desc->data = NULL; desc->len = 0; } @@ -138,21 +136,24 @@ } static int iwl_alloc_fw_desc(struct iwl_drv *drv, struct fw_desc *desc, - struct fw_sec *sec) + struct fw_sec *sec) { - if (!sec || !sec->size) { - desc->v_addr = NULL; + void *data; + + desc->data = NULL; + + if (!sec || !sec->size) return -EINVAL; - } - desc->v_addr = dma_alloc_coherent(drv->trans->dev, sec->size, - &desc->p_addr, GFP_KERNEL); - if (!desc->v_addr) + data = vmalloc(sec->size); + if (!data) return -ENOMEM; desc->len = sec->size; desc->offset = sec->offset; - memcpy(desc->v_addr, sec->data, sec->size); + memcpy(data, sec->data, desc->len); + desc->data = data; + return 0; } diff -urN .iwlwifi/iwl-fw.h iwlwifi/iwl-fw.h --- .iwlwifi/iwl-fw.h 2012-07-21 13:58:29.000000000 -0700 +++ iwlwifi/iwl-fw.h 2012-09-12 16:30:28.997944875 -0700 @@ -124,8 +124,7 @@ /* one for each uCode image (inst/data, init/runtime/wowlan) */ struct fw_desc { - dma_addr_t p_addr; /* hardware address */ - void *v_addr; /* software address */ + const void *data; /* vmalloc'ed data */ u32 len; /* size in bytes */ u32 offset; /* offset in the device */ }; diff -urN .iwlwifi/iwl-trans-pcie.c iwlwifi/iwl-trans-pcie.c --- .iwlwifi/iwl-trans-pcie.c 2012-07-21 13:58:29.000000000 -0700 +++ iwlwifi/iwl-trans-pcie.c 2012-09-12 16:32:48.190750091 -0700 @@ -896,13 +896,10 @@ /* * ucode */ -static int iwl_load_section(struct iwl_trans *trans, u8 section_num, - const struct fw_desc *section) +static int iwl_load_firmware_chunk(struct iwl_trans *trans, u32 dst_addr, + dma_addr_t phy_addr, u32 byte_cnt) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - dma_addr_t phy_addr = section->p_addr; - u32 byte_cnt = section->len; - u32 dst_addr = section->offset; int ret; trans_pcie->ucode_write_complete = false; @@ -915,8 +912,8 @@ FH_SRVC_CHNL_SRAM_ADDR_REG(FH_SRVC_CHNL), dst_addr); iwl_write_direct32(trans, - FH_TFDIB_CTRL0_REG(FH_SRVC_CHNL), - phy_addr & FH_MEM_TFDIB_DRAM_ADDR_LSB_MSK); + FH_TFDIB_CTRL0_REG(FH_SRVC_CHNL), + phy_addr & FH_MEM_TFDIB_DRAM_ADDR_LSB_MSK); iwl_write_direct32(trans, FH_TFDIB_CTRL1_REG(FH_SRVC_CHNL), @@ -935,19 +932,51 @@ FH_TCSR_TX_CONFIG_REG_VAL_DMA_CREDIT_DISABLE | FH_TCSR_TX_CONFIG_REG_VAL_CIRQ_HOST_ENDTFD); - IWL_DEBUG_FW(trans, "[%d] uCode section being loaded...\n", - section_num); ret = wait_event_timeout(trans_pcie->ucode_write_waitq, trans_pcie->ucode_write_complete, 5 * HZ); if (!ret) { - IWL_ERR(trans, "Could not load the [%d] uCode section\n", - section_num); + IWL_ERR(trans, "Failed to load firmware chunk!\n"); return -ETIMEDOUT; } return 0; } +static int iwl_load_section(struct iwl_trans *trans, u8 section_num, + const struct fw_desc *section) +{ + u8 *v_addr; + dma_addr_t p_addr; + u32 offset; + int ret = 0; + + IWL_DEBUG_FW(trans, "[%d] uCode section being loaded...\n", + section_num); + + v_addr = dma_alloc_coherent(trans->dev, PAGE_SIZE, &p_addr, GFP_KERNEL); + if (!v_addr) + return -ENOMEM; + + for (offset = 0; offset < section->len; offset += PAGE_SIZE) { + u32 copy_size; + + copy_size = min_t(u32, PAGE_SIZE, section->len - offset); + + memcpy(v_addr, (u8 *)section->data + offset, copy_size); + ret = iwl_load_firmware_chunk(trans, section->offset + offset, + p_addr, copy_size); + if (ret) { + IWL_ERR(trans, + "Could not load the [%d] uCode section\n", + section_num); + break; + } + } + + dma_free_coherent(trans->dev, PAGE_SIZE, v_addr, p_addr); + return ret; +} + static int iwl_load_given_ucode(struct iwl_trans *trans, const struct fw_img *image) { @@ -955,7 +984,7 @@ int i; for (i = 0; i < IWL_UCODE_SECTION_MAX; i++) { - if (!image->sec[i].p_addr) + if (!image->sec[i].data) break; ret = iwl_load_section(trans, i, &image->sec[i]); diff -urN .iwlwifi/iwl-ucode.c iwlwifi/iwl-ucode.c --- .iwlwifi/iwl-ucode.c 2012-07-21 13:58:29.000000000 -0700 +++ iwlwifi/iwl-ucode.c 2012-09-12 17:10:18.151765877 -0700 @@ -270,7 +270,7 @@ static int iwl_verify_sec_sparse(struct iwl_priv *priv, const struct fw_desc *fw_desc) { - __le32 *image = (__le32 *)fw_desc->v_addr; + __le32 *image = (__le32 *)fw_desc->data; u32 len = fw_desc->len; u32 val; u32 i; @@ -294,7 +294,7 @@ static void iwl_print_mismatch_sec(struct iwl_priv *priv, const struct fw_desc *fw_desc) { - __le32 *image = (__le32 *)fw_desc->v_addr; + __le32 *image = (__le32 *)fw_desc->data; u32 len = fw_desc->len; u32 val; u32 offs; ^ permalink raw reply [flat|nested] 8+ messages in thread
end of thread, other threads:[~2012-09-13 4:52 UTC | newest]
Thread overview: 8+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
[not found] <20120908161038.GF3347@merlins.org>
[not found] ` <1347122927.1234.849.camel@edumazet-glaptop>
2012-09-04 0:30 ` kernel 3.5.2/amd64: iwlwifi 0000:03:00.0: failed to allocate pci memory Marc MERLIN
2012-09-08 17:01 ` Marc MERLIN
2012-09-08 18:57 ` Johannes Berg
2012-09-10 10:24 ` Johannes Berg
2012-09-10 15:12 ` Larry Finger
2012-09-10 15:44 ` Johannes Berg
2012-09-10 15:34 ` Marc MERLIN
[not found] ` <20120910154402.GF5053@merlins.org>
[not found] ` <1347292023.4272.28.camel@jlt4.sipsolutions.net>
2012-09-13 4:52 ` Marc MERLIN
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).